Picture the old way. A soundstage in Culver City, a celebrity with a four-hour window, a crew of forty, a craft services table nobody touches after noon. At the end of the day you walk out with eight, maybe ten usable assets, and everyone acts like that’s a miracle because the talent’s manager was checking a watch the whole time.
Now picture the new way. Same star, same single shoot, except this time they step into a capture rig, get scanned, and walk out. The agency walks out with thirty to fifty.
That’s not a thought experiment. On November 9, Ogilvy and CAA launched Echo, a platform that takes a celebrity’s likeness from The CAA Vault and turns one production shoot into what Ogilvy calls a live, real-time content system. Ogilvy’s own release says a traditional eight-to-10-asset shoot becomes roughly 30 to 50 media-ready assets, potentially cutting average cost per asset by up to 65 to 75 percent. It runs on WPP Open. It’s commercially available across the WPP network this month.
Ogilvy’s global chief innovation officer, Carol Reed, framed it like this: “We are witnessing the ‘NIL moment’ between advertising and entertainment.”
She’s right. Just not about the part everyone’s excited about.
Credit where it’s owed, because the consent architecture here is serious. According to Digiday, CAA clients own their digital likenesses and choose which brands and markets they’ll appear for. Every asset needs talent sign-off before it ships, through an approval portal. Each one carries a watermark that reads “[Talent Name] Approved | Recreated with AI.” The Vault itself, which Digiday reports was built for CAA by Veritone, captures people with 204 cameras. Terry Crews has reportedly opted in. This is the grown-up version of synthetic talent, and it’s a hell of a lot better than the deepfake swamp brands were wading through two years ago.
So we’re not here to clutch pearls about AI. We’re here to point at what just happened to the price of fame.
When a scarce thing becomes abundant, its value doesn’t stay put. It moves. It moves to whatever is still scarce. For a century, celebrity in advertising was expensive because it was rationed: limited days, limited shots, limited usage windows. That rationing was the whole business model. You paid for access, and access was the creative strategy, because if you landed the face, the face did most of the work.
Echo, and everything that will copy Echo within eighteen months, ends the rationing. A Gartner analyst told Digiday it could “fundamentally change celebrity endorsement economics.” Translation: the face is about to stop doing the work.
Here’s the question nobody in the press release asked… What are those fifty assets going to say?
Because volume was never the bottleneck. Nobody in the history of the internet has scrolled past an ad and thought, “If only there had been forty more versions of that.” The bottleneck has always been the thing that makes a person stop: a line they repeat, a joke they send to the group chat, a frame that looks like nothing else in the feed. You can’t scan that into a vault. There’s no 204-camera rig for a point of view.
Fifty versions of a forgettable idea is just a forgettable idea with better distribution.
Think about how you already experience famous people online. You don’t see an actor in one campaign. You see them in a bank ad, a streaming promo, a meme of the bank ad and a fan edit of the meme, all before your coffee gets cold. Ogilvy’s global AI creative lead, David Raichman, pitched Echo as content that’s more plentiful, faster and more trusted. Two out of three is easy. Trust has never come from quantity. It comes from the moment a famous person does something that feels like them, not like a deliverable.
And the audience has already told us they’re not impressed by the trick. Digiday cites YouGov research showing 49 percent of Britons have a negative view of AI-generated actors in ads. Ogilvy’s own commissioned research, to its credit, found that explicit talent approval helps; more than 80 percent of respondents read talent-approved messaging as a cue of sanctioned AI use. Fine. But approval is a permission slip, not a reason to care. A watermark that says the star signed off doesn’t make the work worth watching. It just makes it legal.
Now the turn, the part that should make every brand manager a little uncomfortable.
The real risk of Echo isn’t that it fails. It’s that it works exactly as advertised, and brands do the obvious thing with it. They buy the volume. They feed fifty variants into Advantage+ and Performance Max, let the algorithm sort the winners, and call that creative strategy. Within a year, the feed fills up with the same twelve licensed faces, each one rendered in fifty flavors, each one saying something a committee approved. Celebrity becomes wallpaper. And wallpaper, by definition, is the thing nobody looks at.
We’ve watched this movie before, just with cheaper actors. When stock photography got cheap, every website got a smiling woman with a headset. When UGC got cheap, every brand got a creator holding the product up to a ring light saying “okay so I’ve been using this for two weeks.” Each time, the tool was fine. The thinking was the casualty. Abundance didn’t raise the floor. It flooded it.
That’s the uncomfortable implication. If the face is cheap and the volume is free, the only thing left to compete on is the idea, the cultural read, the specific reason this ad belongs in this feed on this Tuesday. And most marketing organizations have spent the last decade defunding exactly that. They’ve built teams to manage output, not to have opinions. Echo hands those teams a firehose and assumes someone knows where to point it.
There’s a version of this that’s genuinely exciting. Imagine a brand with a real idea, something rooted in a place or a subculture, using a licensed talent likeness to localize it across thirty markets, thirty in-jokes, thirty neighborhoods. A spot that lands differently in Boyle Heights than it does in Brixton, because someone who understands both wrote it. That’s what this technology could be for: one sharp idea, translated by people who know the rooms it’s walking into. Not one shoot, sliced into fifty pieces of nothing.
That’s how we work at BoomHaus, minus the vault. We start with the cultural read, the thing people are actually doing and saying in the places they actually live, and only then decide how much content the idea can carry. Some ideas can carry fifty executions. Most can carry five. A few great ones only need one, posted at the right moment, and the internet does the copying for you for free.
The industry is going to spend the next year bragging about cost-per-asset. It’s a clean metric. It fits on a slide. It makes procurement happy. But cost-per-asset is a measure of what you made, not what anyone felt, and the gap between those two numbers is where brands quietly go broke.
So yes, call it advertising’s NIL moment. College athletes got paid for their names and likenesses, and the market for famous faces got bigger overnight. What it didn’t do was turn a backup point guard into a star. Talent was still talent. Teams that won still won on coaching.
Same rule here. You can license the face. You can clone the voice. You can render fifty takes before lunch.
You still have to give it something to say.
